High Speed Downlink Packet Access

Description

High Speed Downlink Packet Access (HSDPA) is a 3GPP radio interface enhancement for UMTS (Universal Mobile Telecommunications System) that dramatically increases downlink packet data throughput and reduces latency. Architecturally, it introduces new functionalities primarily in the Node B (base station) and the User Equipment (UE), moving key MAC-layer scheduling and retransmission control from the Radio Network Controller (RNC) to the Node B. This reduces processing delays. The key physical channel added is the High-Speed Downlink Shared Channel (HS-DSCH), which is a shared transport channel used to carry user data. It is associated with several downlink and uplink control channels: High-Speed Shared Control Channel (HS-SCCH) for downlink signaling, and High-Speed Dedicated Physical Control Channel (HS-DPCCH) for uplink feedback.

HSDPA operates using several key techniques. Fast Link Adaptation adjusts the modulation and coding scheme (MCS) every 2 ms Transmission Time Interval (TTI) based on channel quality indicator (CQI) reports from the UE. It employs higher-order modulation (16-QAM alongside QPSK) for peak rates. Fast Hybrid Automatic Repeat Request (HARQ) allows for rapid retransmissions at the physical layer, managed by the Node B, which improves reliability and efficiency. Fast Packet Scheduling, also in the Node B, decides which users to serve in each TTI based on channel conditions and fairness algorithms, maximizing cell throughput. The UE uses a buffer to reorder packets received out-of-order due to HARQ processes.

In the network, HSDPA coexists with legacy Dedicated Channel (DCH) services. The RNC retains control for radio resource management, admission control, and mobility management (handovers), but the user plane for HSDPA is routed directly from the Node B. This split architecture allows for a smooth upgrade from Release 99 UMTS. HSDPA was a cornerstone of mobile broadband, enabling peak theoretical rates from 1.8 Mbps in early releases to over 42 Mbps with later multi-carrier and MIMO enhancements. It served as the performance benchmark before the advent of LTE.

Purpose & Motivation

HSDPA was created to address the insufficient data rates and high latency of the initial Release 99 UMTS networks, which were inadequate for emerging internet applications like web browsing with rich content, email with attachments, and early video streaming. The primary problem was the centralized architecture where the RNC handled all scheduling and retransmissions, introducing significant delay (around 80-100 ms RTT) and limiting spectral efficiency and peak user throughput.

The motivation for HSDPA, introduced in Release 5, was to bring internet-like speeds to mobile users and make 3G a competitive broadband technology. It solved the limitations by moving intelligence to the Node B, enabling faster reaction to radio channel variations. Techniques like adaptive modulation, fast scheduling, and HARQ were inspired by concepts from fixed broadband but adapted for the mobile environment. This evolution was driven by operator demand for higher capacity and better user experience to increase data revenue. HSDPA, often termed 3.5G, successfully extended the lifecycle of UMTS networks and paved the way for the packet-optimized design principles later fully realized in LTE.

Key Features

• High-Speed Downlink Shared Channel (HS-DSCH) with 2 ms TTI
• Fast Node B-based packet scheduling and link adaptation
• Support for QPSK and 16-QAM modulation schemes
• Fast Hybrid ARQ (HARQ) with incremental redundancy
• Channel Quality Indicator (CQI) reporting from UE
• Evolution to multi-carrier HSDPA and MIMO in later releases

Evolution Across Releases

Defining Specifications